Fuvgus oj Lolium. 291 



the view that it was a smut. He was drawn to this conclusion by 

 the idea " of the probable progression of the evolution of para- 

 sitism in smuts/' from the loose smut of oats through the loose 

 smut of wheat to the Lolium fungus. The loose smut of wheat, 

 forming as it were, an intermediate stage l^etween the loose smut <; ' 

 oat and the fungus of Lolium. It is well known that the spores c i 

 the loose smut of wheat infect the ovary at the flowering stage, and 

 there form a mycelium, which perennates in the embryo of the 

 grain, growing when it germinat-es in the manner of most smuts, 

 and forming its spores during the next flowering stage. 



The Lolium fungus could be regarded as a development from a 

 type such as this, in which' the sporing stage has been entirely sup- 

 pressed, or at most occurs extremely rarely. 



The points in the life-history of the fungus associated with 

 Lolium pere?me, which could be used to support this view, are : — 

 L The behaviour of the hyphae during the growth of the plant 

 up to the flowering stage, which closely resembles the 

 method adopted by the Ustilagineae. 



2. The formation of knots in the tissue of the young ovary, 

 . which, however, I prefer to explain, not as an attempt 



towards spore-formation, but as the preliminary to cell 

 infection. 



3. The intra-cellular course of the hyphae. 



These, at first sight, undoubtedly seem to weigh heavily as evi- 

 dence in favour of its relation to the smut family, the most sug- 

 gestive being the behaviour of the fungus during the vegetative- 

 period of the grass. 



This mode of growth in the host plant, however, is not limited 

 to the Ustilagijieae. Rayner (20) describes a fungus associated 

 with Galluna vulgaris which grows in the tissues of the plant with- 

 out any external evidence of its presence or without disturbing the- 

 normal groAvth. of tlie host. 



The greatest development of the fungus in this case takes place- 

 on the roots of the plants forming a mycorrhiza, endotrophic in 

 character, but unlike most other mycorrhizal plants, the fungus-- 

 keeps pace with the growing point of the stem, and after the incep- 

 tion of the ovary it enters this organ, and forms a mycelium in 

 the ovary wall. The embryo remains sterile, but infection of the 

 seed-coat is accomplished by the hyphae, so that the production of 

 a mycorrhiza in the roots of the next generation of Calluna is not 

 left to chance. . 



16 



